Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Visual Quality Evaluation on SciVidEval
Loading...
10
VLM-as-Judge Score
Source Paper
2.8864
4.7332
6.58
8.4268
Sep 14, 2025
VLM-as-Judge Score
PPL
Asset Match Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
VLM-as-Judge Score
PPL
Asset Match Accuracy
Source Paper
2025.09
10
-
-
VideoAgent
Backbone=Gemini-2.5 Pr...
2025.09
8.03
21.33
79.24
VideoAgent
Backbone=GPT-4o, Summa...
2025.09
7.7
25.78
73.31
VideoAgent
Backbone=Qwen-2.5VL-7B...
2025.09
7.37
27.92
63.65
AutoSlides
2025.09
6.64
17.31
38.37
PresentAgent
2025.09
5.97
26.64
66.75
Author
2025.09
4.74
19.46
88.62
LunWenShuo
2025.09
3.76
24.08
48.57
NotebookLM
2025.09
3.16
15.37
31.99
Feedback
Search any
task
Search any
task